From 91611cb4ea7b7dfc58b2c4eeef70effd24c9f01b Mon Sep 17 00:00:00 2001 From: Marc Celani Date: Mon, 14 May 2018 16:18:32 -0500 Subject: [PATCH] Add Folly's implementation of T-Digest to README Folly's implementation of t-digests is lightning fast and easy to use. * Easy to use APIs for both single digest aggregation and sliding window aggregation * It's cpu-local buffering enables 60ns writes across threads. * It's merge algorithm leverages std::inplace_merge to speed up digest aggregation (across sliding windows, or when merging cpu-local buffers) * It's smart about it's choice of k function to avoid expensive floating point operations * Well tested, with benchmarks and benchmark results included in the results. * Used in production at Facebook. --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index c48257af..8bd8de2e 100644 --- a/README.md +++ b/README.md @@ -85,7 +85,7 @@ The t-digest algorithm has been ported to other languages: - Python: [tdigest](https://github.com/CamDavidsonPilon/tdigest) - Go: [github.com/spenczar/tdigest](https://github.com/spenczar/tdigest) - Javascript: [tdigest](https://github.com/welch/tdigest) - - C++: [CPP TDigest](https://github.com/gpichot/cpp-tdigest) + - C++: [CPP TDigest](https://github.com/gpichot/cpp-tdigest), [FB's Folly Implementation (high performance)](https://github.com/facebook/folly/blob/master/folly/stats/QuantileEstimator.h) - Scala: need link! Continuous Integration